[00:00:00] Speaker 00:
The next argued case is number 22, 1364, People AI Incorporated against Plan Incorporated.

[00:00:09] Speaker 04:
Good morning.

[00:00:10] Speaker 00:
Mr. Ryanes.

[00:00:11] Speaker 04:
Thank you, Your Honor.

[00:00:12] Speaker 04:
Good morning, and Happy New Year to you.

[00:00:13] Speaker 00:
Thank you.

[00:00:15] Speaker 04:
The patent eligibility decision here is erroneous, and I'd like to identify three principal characteristics of the claimed inventions that demonstrate that.

[00:00:29] Speaker 04:
The technology described in the claims is a technological improvement for a CRM system that creates software to implement objective rules to replace subjective and haphazard decisions that are made by various personnel that the patent itself describes as illustrating infected with bias, subjective bias, and is error prone.

[00:00:59] Speaker 04:
That's the first thing.

[00:01:00] Speaker 04:
The second thing is that these claims, like the claims in Bascom and Finjen, allow for individualized filtering.

[00:01:13] Speaker 04:
What they basically say, and this could get lost.

[00:01:16] Speaker 04:
There are a lot of claims, a lot of words overall.

[00:01:20] Speaker 04:
But basically, the filtering policy that's applied is based on the particular account

[00:01:29] Speaker 04:
And that's right in the claims, 3, 4, 5, and it's elsewhere.

[00:01:34] Speaker 04:
And that individualized filtering where you can have a policy for one account and a different policy for another account is very flexible and by no means something that a post office or office services would be able to do.

[00:01:52] Speaker 04:
The third reason is that the data structures here

[00:01:56] Speaker 04:
or for more efficient management of a CRM system.

[00:01:59] Speaker 04:
The CRM system, typically like a Salesforce system, is a large software system that's dominant in the American industry.

[00:02:08] Speaker 04:
And then there's an industry of companies such as People.ai, a huge innovator, that allow you to process information on a different system.

[00:02:19] Speaker 04:
And that has advantages that are described in the record.

[00:02:22] Speaker 04:
So those are the three main differences

[00:02:25] Speaker 04:
On starting with the first one, the issue, as I read the case law, and as we all know, this area of the case law is not easy for anyone to navigate, but is basically, is this just automation of what's going on?

[00:02:43] Speaker 04:
Is this just automation of what people are doing anyway?

[00:02:47] Speaker 04:
And there is absolutely nothing in this record that shows that anyone

[00:02:53] Speaker 04:
in the office services, in the sales associates, was doing the very specific things that are described in the claims.

[00:03:03] Speaker 04:
And just starting with claim 11 of the 345, particularly graphic, is that you select, based on the electric account, one or more filtering policies.

[00:03:15] Speaker 04:
So you've got a particular account.

[00:03:17] Speaker 04:
And for this person, they don't want x, y, and z particular

[00:03:23] Speaker 04:
email addresses or something, how they've tailored what they don't want included, or what contracts they don't want included, or whatever's tailored to them.

[00:03:31] Speaker 04:
That's a very tailored thing.

[00:03:33] Speaker 04:
And then you might apply a regex policy, if that's what that account wants.

[00:03:38] Speaker 04:
And you might apply a different policy.

[00:03:40] Speaker 04:
Where is that happening in a mailroom?

[00:03:43] Speaker 04:
I mean, it's just not.

[00:03:45] Speaker 04:
I think the better argument from the other side, which is pivoted to, is, well, it's really what a sales account executive might do.

[00:03:55] Speaker 04:
But the reality is the way people would operate in this complex world, this isn't a simple what's spam, knock it out.

[00:04:02] Speaker 04:
This is you've got many projects.

[00:04:04] Speaker 04:
You've got elements of projects.

[00:04:06] Speaker 04:
You've got to figure which pieces you want, which personnel in a large organization, many accounts.

[00:04:12] Speaker 04:
which one goes in which file, that is a judgment call that a salesperson exercises based on the overall, their experience in the matter, and otherwise.

[00:04:23] Speaker 05:
That is a conventional activity I think you would all agree on, right?

[00:04:28] Speaker 05:
The sorting out of

[00:04:30] Speaker 05:
emails or documents and figuring out which file and email or document goes into.

[00:04:39] Speaker 05:
That was a conventional activity.

[00:04:41] Speaker 05:
And I seem to recall that the patent seemed to be saying that manual process can create errors, and it can be very inefficient.

[00:04:51] Speaker 05:
But that's what we're trying to address here.

[00:04:55] Speaker 04:
Right.

[00:04:56] Speaker 04:
And what I would say is, McCrow, which is off-cited, but I think here is really on point, is no different in the sense, and I was involved in that case, so I have more familiarity than I might ordinarily, that people were using morph weight sets, and they were doing it in a subjective, manual way.

[00:05:19] Speaker 04:
So it was conventional in the same way.

[00:05:22] Speaker 04:
I don't think the question is, were people doing

[00:05:25] Speaker 04:
the overall function.

[00:05:27] Speaker 04:
The question is, were people doing the process that's claimed in the patent?

[00:05:33] Speaker 04:
And that's what they weren't doing.

[00:05:35] Speaker 04:
What they were doing is they were making a subjective call.

[00:05:37] Speaker 04:
It refers to human bias in column one.

[00:05:42] Speaker 04:
They were making a subjective call based on whatever that person thought at the time.

[00:05:47] Speaker 04:
And another person in another department or another contract, another team thought something else.

[00:05:51] Speaker 04:
And they were making a Gestalt-based call as to, well, this looks like it's related to this, or this looks like this shouldn't go to x and y. Instead of a systematic way that, just like McRoe, where once you have a technological problem, you can solve it.

[00:06:12] Speaker 04:
Now look, there's probably costs.

[00:06:14] Speaker 04:
to having it automated with this way, where you have these particular rules, in the sense that maybe sometimes a human could do a better job.

[00:06:25] Speaker 04:
But the main point is that it's a different job.

[00:06:28] Speaker 04:
What the human is doing is different than having node profiles, selecting a policy based on one of three things, applying one of three policies.

[00:06:37] Speaker 04:
I mean, that's not the real world.

[00:06:39] Speaker 05:
I'm trying to figure out, couldn't a human

[00:06:42] Speaker 05:
person apply these same filtering rules that are recited in the claim?

[00:06:49] Speaker 05:
Like pick up a given email, read it, and see whether or not it has, I don't know, human resources content, or has somebody's social security information and say, well, this doesn't apply to me, or I'm going to filter out this email.

[00:07:07] Speaker 05:
I mean, that's a good question, of course.

[00:07:11] Speaker 05:
I have to use some dangerous language for you.

[00:07:14] Speaker 05:
Isn't that an abstract mental process?

[00:07:17] Speaker 04:
That's to say, again, that would apply in Finjin and Bascom and all those cases in the sense of, can you take the process

[00:07:33] Speaker 04:
that's in the claim and implement it by humans if you really wanted to.

[00:07:38] Speaker 04:
If you put enough effort into it with enough humans, with enough note cards as to what they're supposed to do, could you ultimately accomplish that?

[00:07:46] Speaker 04:
And that's sort of backwards, right?

[00:07:48] Speaker 04:
That's not conventional.

[00:07:49] Speaker 04:
That's what could be in the future.

[00:07:52] Speaker 04:
So to me, it's sort of a backwards question.

[00:07:54] Speaker 04:
The question is, were people doing it?

[00:07:56] Speaker 04:
And the record is zilcho on people doing the selecting policies based on individual accounts, having no profiles that are outside of the CRM.

[00:08:10] Speaker 01:
Council, just let me understand.

[00:08:11] Speaker 01:
Appendix page 1067, right?

[00:08:14] Speaker 01:
Do you have the appendix with you?

[00:08:18] Speaker 01:
Yes.

[00:08:18] Speaker ?:
Excellent.

[00:08:19] Speaker 01:
So this person, Mr. Silver, is this person the person who was arguing on behalf of your client, if I'm not mistaken?

[00:08:29] Speaker 04:
Do you know?

[00:08:34] Speaker 04:
Be honest with you.

[00:08:35] Speaker 04:
I'm not sure I know by the name.

[00:08:38] Speaker 04:
But let me look at the.

[00:08:42] Speaker 01:
So once you get to there, I was just trying to follow up on this conventional activity point.

[00:08:46] Speaker 01:
It looked like they were basically acknowledging that this would be conventional activity and really talking about whether or not you're trying to automate an abstract idea.

[00:08:58] Speaker 04:
If you're referring to line 6 through 12, and I'm a little on the fly here, so apologies this, as a counsel, the argument is that the patents.

[00:09:07] Speaker 04:
So that's just reciting what the argument is.

[00:09:11] Speaker 04:
they conceded they were conventional because it's not true, and it would have been against interest.

[00:09:15] Speaker 04:
So there's two reasons why I don't think that would have been.

[00:09:18] Speaker 04:
But if you're referring to what's in line eight, it says the argument is that.

[00:09:24] Speaker 05:
But what about further down on the same page, like starting at line 14, for example?

[00:09:30] Speaker 05:
So I go on and I drag my emails in, and I'm going to manually do that as a person.

[00:09:35] Speaker 05:
That's the conventional activity the patent talks about.

[00:09:41] Speaker 05:
and then further down at line 19.

[00:09:43] Speaker 05:
Now the problem that arises is when you try to automate this abstract idea, when you try to, you know, programatize it, use a computer to do it, you run into significant problems and a big problem you run into is this problem of the API calls.

[00:09:59] Speaker 05:
And so as

[00:10:01] Speaker 05:
I guess following up on what Judge Cunningham was inquiring about, this page from the hearing below seems to suggest that, OK, doing all of this sorting of emails and maybe dragging emails over into different files, that's an abstract idea.

[00:10:21] Speaker 05:
But once you try to programmatize it, automate it, now you're going to run into a technical problem

[00:10:29] Speaker 05:
the act of attempting to automate this abstract idea.

[00:10:33] Speaker 05:
So in the end, it appears there's a fair way of reading this as saying, OK, this activity of sorting emails is an abstract idea, but to automate it, we contend is that what we did was a technical solution to a technical problem.

[00:10:51] Speaker 05:
Is that a fair understanding of this page?

[00:10:54] Speaker 04:
I don't think so, and let me explain why.

[00:10:58] Speaker 04:
I think this underlines the themes that I've enunciated to start the argument.

[00:11:06] Speaker 04:
We're not in a mail room.

[00:11:09] Speaker 04:
We're at a CRM system using computers.

[00:11:12] Speaker 04:
So the conventional activity isn't anything that the district court was talking about or, frankly, opposing counsel in their briefs about, well, what would happen if someone was doing this?

[00:11:23] Speaker 04:
What we have is the way it was done was people, I mean, it couldn't be more haphazard, dragging emails as they see fit.

[00:11:33] Speaker 04:
onto different files.

[00:11:35] Speaker 04:
Now, I don't know about you.

[00:11:37] Speaker 04:
When you drag them, I drag them.

[00:11:38] Speaker 04:
I'm not applying rules or doing any of the things that are being described in terms of a node profile and transmitting into the CRM.

[00:11:50] Speaker 01:
The claims are not limited to CRMs, right?

[00:11:52] Speaker 01:
They also are really focused on system of record, which is broader, correct?

[00:11:56] Speaker 04:
That is correct.

[00:11:57] Speaker 04:
I think for what the parties have stated, and I think you'll find in there,

[00:12:02] Speaker 04:
It's essentially a CRM, but you're right that the specification does include a series of items of which the CRM is the most conspicuous.

[00:12:11] Speaker 04:
I think there's some human resource files where it might system.

[00:12:15] Speaker 04:
But the point is that where the objects are,

[00:12:22] Speaker 04:
in the system of record is different from where you're processing information such as no profile in these relationships.

[00:12:31] Speaker 04:
But I do want to go back briefly, which is to say what was conventional was people were dragging and dropping.

[00:12:37] Speaker 04:
That's not telling us anything that they're not doing wasn't subjective.

[00:12:42] Speaker 04:
And I think this judgment on the pleadings at this early stage where

[00:12:50] Speaker 04:
there's no record that people were doing any of the structural things or the decision-making processes in the prior art makes this case very much like the kinds of cases where, like Bascom, where it's like, on this record, with inferences going for the patiny, we're not going to say that this is just implementing that which was done manually.

[00:13:12] Speaker 04:
All that this page says, and it's a good question, is that people were dragging and dropping, and then you have problems.

[00:13:19] Speaker 04:
And the problems are the API, which is that you only have a certain amount of calls.

[00:13:23] Speaker 04:
So then you have to create new structures.

[00:13:25] Speaker 04:
And the new structures are node profiles and other ways of processing information outside of the system of record, which is typically in CRM.

[00:13:34] Speaker 04:
So you have that structural point.

[00:13:36] Speaker 04:
And I'd be interested what the answer is on the individualization of it.

[00:13:41] Speaker 04:
The fact that you can have a different policy set,

[00:13:47] Speaker 04:
regex or logic or different things based on each individual.

[00:13:53] Speaker 04:
No one has shown that was in the prior art.

[00:13:55] Speaker 04:
No one has shown that's conventional.

[00:13:57] Speaker 04:
Let me wrap up at this point to say, the problem with this area of the law, as we all know, it's a challenging one.

[00:14:04] Speaker 04:
It's not easy.

[00:14:05] Speaker 04:
is this sort of like, well, people knew how to do all this.

[00:14:09] Speaker 04:
And so you're just putting in software what everybody knew.

[00:14:11] Speaker 04:
And it's at the outset of the case.

[00:14:14] Speaker 04:
And people don't even know what a node profile.

[00:14:17] Speaker 04:
And a lot of the claim instructions seem sloppy.

[00:14:19] Speaker 00:
We're starting to explore it.

[00:14:21] Speaker 00:
I'm having trouble in deciding if there's any way of being more precise about the boundary line between section 101 and, let's say, 103.

[00:14:32] Speaker 00:
And the arguments that I've been hearing

[00:14:34] Speaker 00:
could apply just as well to 103 as to 101.

[00:14:39] Speaker 00:
But is it correct that all that was considered at this stage in the litigation was 101?

[00:14:47] Speaker 04:
That's correct.

[00:14:48] Speaker 04:
And the rigor, if you're going to engage in this, well, this was all known, and you're just putting it on a computer, if we're going to do that, it needs to have the rigor and protections of Section 103.

[00:14:57] Speaker 01:
Let's have a couple of housekeeping questions.

[00:14:58] Speaker 01:
Of course, obviously.

[00:14:59] Speaker 04:
Yes, thank you.

[00:15:00] Speaker 01:
Can we focus our analysis on the same claims the district court focused on?

[00:15:04] Speaker 04:
Um, yes, that's fine.

[00:15:07] Speaker 01:
Don't need to address any others, right?

[00:15:09] Speaker 04:
Yes, Your Honor.

[00:15:11] Speaker 01:
And then my other housekeeping question is, is there anything in the claims that says that node profiles have to include associations between profiles?

[00:15:20] Speaker 04:
You know, I was preparing for that question.

[00:15:22] Speaker 04:
And I believe that the specification supports that.

[00:15:26] Speaker 04:
And I don't believe it was contested.

[00:15:29] Speaker 04:
And that's the whole point of the node profiles.

[00:15:32] Speaker 04:
But this goes to the prematurity point that I was making, which is,

[00:15:36] Speaker 04:
it just seems to me he's coming in at the appellate level that this is like completely under process to be throwing this out based on someone thinking they know exactly what a node profile is or that what the relationship between the system of record

[00:15:48] Speaker 04:
And the data structures such as node profile, I just don't think that's been done in a way that would permit a pleading to throw it out.

[00:15:59] Speaker 05:
Could I think of node profiles as like a Rolodex, electronic version of a Rolodex, or at least node profiles encompasses

[00:16:07] Speaker 05:
electronic Rolodex where you got all the personal contacts every all these different individual names and then their associated phone numbers email addresses and any other personal information with two exceptions so it's not crazy that there are some element of that one is that

[00:16:26] Speaker 04:
Because this is a technological solution to a technological problem.

[00:16:30] Speaker 04:
This isn't some broad-brush thing.

[00:16:31] Speaker 04:
This is for CRM type systems.

[00:16:34] Speaker 04:
Having them outside of the system where they're in their own data structures that can then be batched over and transmitted, which is what claims require here, makes it different.

[00:16:46] Speaker 04:
Like a Rolodex just doesn't have a meaning like that.

[00:16:50] Speaker 04:
To me, to say that's separate from some other thing doesn't make sense.

[00:16:53] Speaker 04:
The second thing is Judge Cunningham's good question, which was, are they associated?

[00:16:59] Speaker 04:
And yes, the concept of the no profiles is that they're linked together.

[00:17:05] Speaker 04:
associated, and that you can build up this rich environment, which I don't think applies to Rolodex.

[00:17:11] Speaker 04:
But if you had to follow up on that.

[00:17:12] Speaker 05:
And finally, any of the representative claims we've been looking at, do any of them require more than looking at two emails?

[00:17:20] Speaker 05:
I mean, some of these claims talk about a first email, a second email, and then some of the others.

[00:17:26] Speaker 05:
Electronic activity, excuse me.

[00:17:28] Speaker 05:
And then some of these other claims just talk about, all right, let's

[00:17:32] Speaker 04:
I want to double check, but I think you're right that it's electronic activity, so it's far more than just email.

[00:17:37] Speaker 04:
But yeah, I think you're right that it's two.

[00:17:40] Speaker 04:
But to the extent that's suggesting that there isn't individualization or customization, I disagree with that, because it clearly says the policy is based on plus you have two, and that's enough to have that kind of distinction.

[00:17:53] Speaker 04:
Judge Cunningham, do you have any house to keep him or otherwise?

[00:17:56] Speaker 04:
Anything else?

[00:17:57] Speaker 00:
No, no.

[00:17:57] Speaker 00:
I'm going to stop.

[00:17:58] Speaker 04:
OK.

[00:17:58] Speaker 04:
All right.

[00:17:58] Speaker 04:
Well, thank you all very much.

[00:18:00] Speaker 00:
We'll save you rebuttal time.

[00:18:02] Speaker 00:
Let's hear from the other side.

[00:18:05] Speaker 00:
Mr. Novikov.

[00:18:06] Speaker 03:
Good morning.

[00:18:07] Speaker 03:
Gene Novikov for Appellee Clary.

[00:18:10] Speaker 03:
I'd like to actually start by addressing the question that Judge Cunningham asked toward the end of Mr. Reinus' argument, which is, do the claims require associations between node profiles?

[00:18:23] Speaker 03:
And Mr. Reinus gave an answer that I thought was a little bit curious in view of the fact that below,

[00:18:34] Speaker 03:
The appellant People.ai wrote, People.ai is stating that a node graph, which is this network of associations that Judge Cunningham was referencing, is not a requirement in the claim.

[00:18:47] Speaker 03:
That's on appendix 5551.

[00:18:49] Speaker 03:
And I'm not sure how that can be much clearer.

[00:18:55] Speaker 03:
Going back to the sort of overall gestalt here, this court held in the Samantha case that, quote, it was long prevalent practice for people receiving paper mail to look at an envelope and discard certain letters

[00:19:13] Speaker 03:
without opening them from sources from which they did not wish to receive mail based on characteristics of the mail.

[00:19:20] Speaker 03:
The list of relevant characteristics could be kept in a person's head.

[00:19:23] Speaker 03:
Characterizing email based on a known list of identifiers is no less abstract.

[00:19:31] Speaker 03:
What these claims add to that description of the claims in the Samantha case is trivial at best.

[00:19:39] Speaker 03:
They say the Samantha claim said you can look at the sources and use them to discard mail that you may consider to be junk or spam.

[00:19:49] Speaker 03:
These claims say, well, you can look at the sources

[00:19:52] Speaker 03:
Or you can look at keywords in the message, or you can look for patterns in the message like social security numbers or credit card numbers or things like this.

[00:20:03] Speaker 03:
And then once you've done that, you could look at the sender and the recipient and you can file the email.

[00:20:09] Speaker 03:
Under this court's decision in fair warning IP, asking those sorts of questions, which are questions that salespeople dragging emails to folders, which is admitted to have been the conventional practice, precisely the questions that those salespeople would have asked.

[00:20:29] Speaker 03:
They would have looked at, what is this email about?

[00:20:31] Speaker 03:
Do I recognize a word in it?

[00:20:33] Speaker 03:
They would have filed the email based on that.

[00:20:36] Speaker 03:
Who is this email from?

[00:20:38] Speaker 03:
What company file or deal was it associated with?

[00:20:43] Speaker 03:
File the email based on that.

[00:20:45] Speaker 03:
If it has somebody's credit card number or something like that, you flag it for special treatment as containing.

[00:20:52] Speaker 05:
Do you have to win on this prior art theme of whether these particular filtering policies were known in the prior art and were used before in order to get an affirm here?

[00:21:06] Speaker 05:
Because in the end, these are all part and parcel of some abstract mental process, arguably.

[00:21:14] Speaker 03:
I think that's right, Your Honor.

[00:21:16] Speaker 03:
I think that this court has had no hesitation ruling in cases like Symantec and cases like content extraction that things like this are a long practiced economic activity, business practice.

[00:21:32] Speaker 03:
And you do not require evidence in the record that there were specific people doing these specific things.

[00:21:40] Speaker 05:
Right.

[00:21:41] Speaker 05:
That's the point I'm trying to get at.

[00:21:44] Speaker 05:
the claim is doing is automating something that humans could do, that humans could practice on their own, regardless of whether humans had in fact practiced that set of thinking steps before in the art.

[00:22:01] Speaker 05:
Doesn't matter, because in the end, what we're dealing with, we've translated that kind of analysis, mental analysis, as being an abstract idea.

[00:22:14] Speaker 03:
I think that's exactly right, and that's exactly the reasoning that the court engaged in in the excerpt that I read from the intellectual ventures to be semantic case.

[00:22:25] Speaker 03:
I think that's exactly right.

[00:22:26] Speaker 03:
And I really want to emphasize Mr. Reines gave individual filtering, individualized filtering as an example of something that was

[00:22:36] Speaker 03:
complicated or technological and not something that a salesperson would have done.

[00:22:42] Speaker 03:
There's only one claim that's addressed in the briefs that has this notion of selecting a filtering policy based on the electronic account.

[00:22:52] Speaker 03:
It's claim 11 of the 345 patent.

[00:22:54] Speaker 03:
And all it says, and I'm not, I would welcome Mr. Reines to correct me.

[00:22:59] Speaker 03:
I don't think I'm oversimplifying at all.

[00:23:01] Speaker 03:
All it says is you look at whose correspondence you're

[00:23:06] Speaker 03:
working with, and you apply different rules based on who that person is.

[00:23:11] Speaker 03:
And that's it.

[00:23:12] Speaker 03:
So that limitation would 100% be satisfied by a salesperson applying the rule that an email from my wife

[00:23:23] Speaker 05:
Shouldn't be filed in a company file because that's a rule that's specific to him and it's something that he's You're right that example sounds a tad trivial but I guess Mr. Reines's larger point is that what the claimed invention was perhaps directed to is this customization of a filtering process and so we have this custom tailored filter that creates a lot of flexibility and

[00:23:53] Speaker 05:
because now across a whole set of individuals, you can have a customized filtering going on.

[00:24:02] Speaker 05:
And so in that sense, does that get him closer to something that could be argued as improving the functionality of the computer as opposed to merely running an abstract idea on a computer?

[00:24:16] Speaker 03:
I don't think it does in this case.

[00:24:17] Speaker 03:
I think it would certainly theoretically be possible to patent and claim a method of customization that enables

[00:24:30] Speaker 03:
implementing it in a large corporate structure.

[00:24:33] Speaker 03:
I actually think the Finjen case is a good example of that because in Finjen, what you had was an exploit of the way that the internet works and the way that ISBs work.

[00:24:46] Speaker 03:
I'm sorry, I'm thinking of Bascom, I apologize.

[00:24:50] Speaker 03:
That's right, Bascom is a good example of that because you're exploiting the way that the internet works and the way that ISPs work and you're saying well if you put the filter at the ISP, the ISP can tell who is making the request

[00:25:04] Speaker 03:
and can implement personalized filtering policies without letting the teenager on the computer mess with the filters.

[00:25:11] Speaker 03:
You don't have that here.

[00:25:12] Speaker 03:
You here just have the statement, customize it.

[00:25:15] Speaker 03:
Based on the electronic account, select one or more filtering policies associated with the data source provider.

[00:25:23] Speaker 03:
There was a reference to the selection of keywords or regular expressions or

[00:25:29] Speaker 03:
a logic-based policy.

[00:25:31] Speaker 03:
But again, it's disjunctive.

[00:25:34] Speaker 03:
That would just be satisfied by selecting the keyword policy and having the salesperson look for his wife's name in the mail.

[00:25:46] Speaker 01:
I just want to close the loop on something that you brought my attention to.

[00:25:50] Speaker 01:
If you look at appendix 5551 to 5552, you were talking about no graphs.

[00:25:57] Speaker 01:
That's right.

[00:25:59] Speaker 01:
When I turned to the top of 5552, it seems to clarify that people.ai is stating that a note graph is not a requirement in the claim, but it's not arguing that notes do not need to be associated with one another.

[00:26:15] Speaker 01:
So I guess what I'm trying to make sure I understand correctly is it seems like note associations is something different than note graphs.

[00:26:21] Speaker 01:
I thought that you were implying that they were the same when I heard you speak earlier.

[00:26:26] Speaker 01:
I just want to make sure I understand.

[00:26:29] Speaker 03:
I think that I may let Mr. Weinberg address this at the risk of having sandbag him, because this is from his matter.

[00:26:40] Speaker 03:
But I think that where it ended below is that people.ai conceded that they needed to plead some sort of association between two records.

[00:26:56] Speaker 03:
which is to say, like, if you had a record for a person and there was a company name in there, and then you had a separate record from a company, that qualified as an association.

[00:27:09] Speaker 03:
So I think to go that far, they can see that that was required.

[00:27:14] Speaker 03:
But this notion of a node graph, whereas Mr. Reinus was saying you had a network of different profiles connected in different ways, and that could be strengthened

[00:27:26] Speaker 03:
in various dynamic and intelligent ways as you build up more information.

[00:27:30] Speaker 03:
That, I think, unequivocally got this claim.

[00:27:33] Speaker 01:
And you agree that the only claims we need to address are the ones that the district court focused on?

[00:27:38] Speaker 03:
I do agree.

[00:27:40] Speaker 03:
And I will, if the court has no more questions, cede the rest of my time to my colleague.

[00:27:46] Speaker 00:
OK.

[00:27:46] Speaker 00:
Let's hear from Mr. Weinberg.

[00:27:49] Speaker 02:
Thank you very much, Your Honor.

[00:27:50] Speaker 02:
Good morning.

[00:27:50] Speaker 02:
Jonathan Weinberg first at sale.

[00:27:52] Speaker 02:
May it please the court.

[00:27:54] Speaker 02:
We've covered a lot of ground and I'm batting clean up here so I would love to address a number of issues but more than anything I'd love to address any issues that your honors have that's outstanding and perhaps that's the node profile issue.

[00:28:05] Speaker 01:
He suggested that you were going to say something about it.

[00:28:08] Speaker 01:
So I'm expecting you to say something about it.

[00:28:09] Speaker 01:
So go for it.

[00:28:10] Speaker 02:
Absolutely.

[00:28:11] Speaker 02:
The node profile was something that was litigated for more than a year.

[00:28:15] Speaker 02:
So I kind of want to push back a bit.

[00:28:18] Speaker 02:
Mr. Reinus's suggestion that this is done on the pleadings and nobody knows what's going on here.

[00:28:24] Speaker 02:
There was initially a motion to dismiss.

[00:28:26] Speaker 02:
We said, what is a node profile?

[00:28:28] Speaker 02:
They said, oh, it's just a data structure that corresponds to a person.

[00:28:33] Speaker 02:
And we said, OK, well, we don't think that that's very specific.

[00:28:39] Speaker 02:
The case was dismissed because they didn't have a node graph suggested.

[00:28:44] Speaker 02:
Following that, they said, OK, a node graph is not really required.

[00:28:47] Speaker 02:
That's how they got the move for leave to amend to file a new complaint, and they said a node graph is not required.

[00:28:52] Speaker 02:
On the back of that, the district court allowed them to file the second amended complaint.

[00:28:56] Speaker 02:
Then they went all the way to summary judgment briefing, and they proposed a node profile

[00:29:03] Speaker 02:
claim construction that just said that it is a data structure corresponding to a person.

[00:29:10] Speaker 02:
And what they accused in the set sale system was that the set sale system had a data structure that was a contact.

[00:29:17] Speaker 02:
And they said, look, here is a person, so contact.

[00:29:20] Speaker 02:
That's a node profile.

[00:29:22] Speaker 02:
And there are connections between the node profiles because a profile has the name of the company where that person works at.

[00:29:31] Speaker 02:
And if it has the name of the company, well, then that's kind of pointing to another profile.

[00:29:36] Speaker 02:
Bob works at Coca-Cola.

[00:29:37] Speaker 02:
Therefore, Bob is associated with Coca-Cola.

[00:29:40] Speaker 02:
So kind of bringing this back down to reality, to what kind of technologies were actually being accused here.

[00:29:46] Speaker 02:
They're really very basic things.

[00:29:49] Speaker 02:
It's really easy to decide this case if we just keep in mind what the prior art was, the admitted system.

[00:30:00] Speaker 02:
Inventors claimed that they invented here.

[00:30:02] Speaker 02:
And that is, they admit that there was a salesperson who received an email from Bob at Coca-Cola.

[00:30:08] Speaker 02:
That salesperson thought in their head, I know who Bob is.

[00:30:11] Speaker 02:
I associate him with a profile in my head.

[00:30:13] Speaker 02:
I know who Coca-Cola is.

[00:30:15] Speaker 02:
I associate that with a profile in my head.

[00:30:17] Speaker 02:
I know that this email is probably concerning this deal that I'm working on with Bob about selling a bunch of bottles to Coca-Cola.

[00:30:24] Speaker 02:
So I know what this email concerns based on who sent the email.

[00:30:29] Speaker 02:
And all of these functions and inventive concepts, they all exist in exactly that prior art system.

[00:30:39] Speaker 02:
The idea here being we're going to replace the salesperson with a computer that does all these same things automatically.

[00:30:46] Speaker 02:
And there's nothing new.

[00:30:47] Speaker 02:
We ask, for example, the email comes in, and it's from the salesperson's wife.

[00:30:53] Speaker 02:
He knows it's personal.

[00:30:54] Speaker 02:
So he sees his wife's email address.

[00:30:56] Speaker 02:
He doesn't put it into the CRM database.

[00:30:59] Speaker 02:
So what's so interesting about a computer doing the same thing?

[00:31:02] Speaker 00:
Remember, the rules are a lot... That's not really... This is a Section 101 action.

[00:31:08] Speaker 02:
Well, the section... Isn't it?

[00:31:10] Speaker 02:
The Section 101 question under McCrow is, is the computer operating in a way that is foundationally fundamentally different than how a person would operate?

[00:31:19] Speaker 02:
In this case, the answer is no, because the person sees the email from his wife and decides not to put it into the CRM.

[00:31:26] Speaker 02:
The computer sees that the email is from a specific email address and decides not to put it into the CRM.

[00:31:32] Speaker 02:
It's the same rule.

[00:31:35] Speaker 02:
As for the user tailored rules, just one comment on that.

[00:31:39] Speaker 02:
That doesn't appear in the 229 and 129 patents, which are asserted against set sale.

[00:31:44] Speaker 02:
There is a contention that they appear there.

[00:31:47] Speaker 02:
I don't see where that is.

[00:31:48] Speaker 02:
So that particular argument doesn't really fly for set sale.

[00:31:53] Speaker 02:
But I will say that in the analog context, of course every person would apply their own rules, wouldn't they?

[00:32:02] Speaker 02:
one person's wife comes in, as opposed to somebody else's wife, there's going to be different rules, right?

[00:32:07] Speaker 02:
So I don't know what's so specific about having the filtering being done, having different policies for different people.

[00:32:15] Speaker 02:
Are there any other questions about the node profiles or ATSA's theorem stuff?

[00:32:23] Speaker 02:
They did make a big deal at the end about the

[00:32:27] Speaker 02:
the data being analyzed outside the CRM.

[00:32:29] Speaker 02:
But we see that this is, again, what the salesperson was doing, analyzing the data in their head, associating Bob with the profile, and deciding whether or not to.

[00:32:39] Speaker 02:
to store that in the CRN.

[00:32:40] Speaker 05:
I think the argument goes that this is a technical improvement because now you can do batch updates without having to rely on the much larger load of API calls.

[00:32:51] Speaker 02:
In order to make that argument, they would have to point to something in the spec or in the claims or anywhere that says that prior art systems didn't do it this way, didn't have the capability to do batch updates.

[00:33:03] Speaker 02:
That was the distinction between University of Florida and Bascom.

[00:33:06] Speaker 02:
You see at the end of University of Florida, that's exactly the distinction they made.

[00:33:09] Speaker 00:
It's not whether the prior art systems do it this way.

[00:33:14] Speaker 00:
That's 103.

[00:33:15] Speaker 00:
The question is, do the prior art systems do it at all?

[00:33:20] Speaker 02:
Right.

[00:33:20] Speaker 02:
Well, the question is, is that described as an inventive concept anywhere in the record?

[00:33:26] Speaker 02:
Does the patent itself say, previous systems did it this way.

[00:33:30] Speaker 02:
We did it a new way.

[00:33:31] Speaker 02:
We invented this.

[00:33:31] Speaker 02:
It was beneficial.

[00:33:33] Speaker 02:
It created fewer API calls.

[00:33:35] Speaker 02:
If there's nothing like that in the record, there's nothing like that in the spec.

[00:33:40] Speaker 02:
It's hundreds of pages of spec.

[00:33:41] Speaker 02:
There's nothing like that in any of the pleadings.

[00:33:44] Speaker 02:
They spent 60 paragraphs in the complaint talking about why this is 101 eligible.

[00:33:48] Speaker 02:
They didn't mention this concept once.

[00:33:50] Speaker 02:
They came up with it at the 101 stage, and that's where you see Mr. Silva going straight to it.

[00:33:56] Speaker 02:
I see that my time is up, but I'm happy to answer any more questions if you have them.

[00:33:59] Speaker 00:
Any more questions?

[00:34:04] Speaker 00:
Thank you, Your Honor.

[00:34:06] Speaker 04:
So the statement that I heard that got the biggest reaction was, you can just argue what's long practice and you don't need any proof.

[00:34:19] Speaker 04:
And then what was argued is the key issue is whether what is being automated in the software system, technologically, is fundamentally different from how people worked.

[00:34:33] Speaker 04:
but that we don't need to put evidence in of what people actually did.

[00:34:37] Speaker 04:
How could that be right?

[00:34:38] Speaker 04:
How could it be right that we don't need to know what actually went before to determine if this is just automating that which happened?

[00:34:46] Speaker 04:
The Coca-Cola example is just, I mean, that I don't think maps to what wouldn't meet the claims, for one thing.

[00:34:53] Speaker 04:
And two, it's laughable how it characterizes the claims.

[00:34:57] Speaker 04:
And two is there's no evidence in the record that's what they did.

[00:35:00] Speaker 04:
We know they drag dropped.

[00:35:01] Speaker 04:
We know there was human bias in doing it.

[00:35:04] Speaker 04:
It was error prone.

[00:35:05] Speaker 04:
because everyone has a different impression of what might be related to which project or all the variety of things this covers.

[00:35:13] Speaker 04:
But the main point, and this goes to Judge Shen, what seems to be your concern is, well, if you can take

[00:35:19] Speaker 04:
or 354 patent, 345 patent, claim 11, and go out and get 100 workers to do it on pieces of paper, then it's invalid, which is like reverse 103.

[00:35:30] Speaker 04:
It's even worse than doing 103 in 101.

[00:35:32] Speaker 04:
It's reverse 103.

[00:35:33] Speaker 04:
But that breaks down, Your Honor, that maybe that passes for step one in a rough justice world of 101.

[00:35:41] Speaker 04:
But it can't be right for step two.

[00:35:43] Speaker 04:
for step two to show that doing all of the things that are in these claims with node profiles and doing out of the system, where you transmit to the system of record what you figured out, and you've selected each of them based on the account, and you've done all of those things, that that was conventional.

[00:36:04] Speaker 04:
That couldn't possibly be inventive to put these series of things together, because

[00:36:11] Speaker 04:
It's long been practiced, but we don't need to put evidence in.

[00:36:16] Speaker 04:
That can't be the standard.

[00:36:17] Speaker 04:
That can't be the way we do this.

[00:36:19] Speaker 04:
And then to have the insult to injury is the end of the argument is we're required.

[00:36:25] Speaker 04:
The burden's on them, clearing convincing evidence.

[00:36:28] Speaker 04:
We're required to come forward and show that the prior art doesn't have certain things, which is really turning things upside down.

[00:36:36] Speaker 04:
I think this area of the law is already the Wild West.

[00:36:40] Speaker 04:
And to do this on a pleadings motion where we don't even know whether there needs to be associations in the no profile.

[00:36:46] Speaker 04:
We have counsel standing up saying it's not a no profile network, which is a different thing than whether there's associations, which we all know that there now is a concession that associations are required.

[00:36:59] Speaker 04:
So the judge was proceeding on the concept that that wasn't required.

[00:37:04] Speaker 04:
So what we need to do is do claim construction.

[00:37:08] Speaker 04:
This is one claim that there were summary judgment papers filed.

[00:37:12] Speaker 04:
There wasn't full fact discovery for the case.

[00:37:14] Speaker 04:
There wasn't decisions on any of the summary judgment.

[00:37:17] Speaker 04:
We need a claim construction proceeding on these relevant concepts, and then develop what's actually being argued as the baseline of what's been automated, rather than just saying, it's been long practiced.

[00:37:29] Speaker 04:
Good enough.

[00:37:30] Speaker 04:
Thank you.

[00:37:31] Speaker 04:
Unless there's any questions.

[00:37:37] Speaker ?:
That concludes this panel's arguments for today.

[00:37:40] Speaker ?:
Thank you.